ClawHub
Back to skill

Security audit

Composite Scene

Security checks across malware telemetry and agentic risk

Overview

This skill provides clear image-compositing instructions, with its external image-model use and multi-image inputs disclosed and aligned with its purpose.

Before installing, understand that this skill is meant for cases where you provide multiple images to be fused into one scene. Do not use it for sensitive private images unless you are comfortable sending those images to the configured image-generation provider.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (1)

Vague Triggers

Medium
Confidence
84% confidence
Finding
The trigger examples are broad, conversational phrases such as 'combine these two photos' and 'make these into one image', which can overlap with many ordinary image-editing requests. This can cause unintended invocation of the skill in contexts where a more specific or safer skill should be selected, leading to incorrect tool routing and unintended image operations.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.